The evaluation of expression levels of the estrogen receptor (ER) and progesterone receptor (PR) in endometrial cancer is a very common procedure in modern practice. However, the significance of such tests remains controversial, and the evaluation of the status of steroid receptors seems to be a more thoroughly justified practice. The present study was carried out with 145 endometrial cancer patients, all of whom had undergone operations at the Seoul National University Hospital, from 1993 to 2002. Paraffin-embedded tissue blocks were sectioned and immunostained with monoclonal anti-ER and anti-PR antibodies. Clinicopathological variables were also analyzed, with 10% cutoff values for ER and PR positivity. A multivariate Cox regression analysis was performed in order to estimate the influences of several clinicopathological and immunohistochemical covariates on patient survival. Forty seven specimens (32.4%) stained as ER positive, and 110 (75.9%) stained as PR-positive. Patients with PR-positive tumor tended to be both younger and more obese than patients with PR-negative tumors (P=0.020, 0.016). Well-differentiated tumors were found to be positive for ER or PR more frequently (P=0.015, <0.001). ER or PR-positive tumors exhibited significantly less myometrial invasion (P=0.042, 0.002). However, multivariate Cox regression analyses revealed that the expressions of ER and PR were not significantly associated with patient survival, and only advanced FIGO stage constituted an independent prognostic factor. Our results suggested that the evaluation of steroid receptors might prove helpful prior to surgery. Low ER and high PR values suggest that racial differences and/or sequential loss of receptors during carcinogenesis might be involved in the expression of steroid receptors in cases of endometrial cancer.